Dušan Arsovski

Singer on Killing Animals

In his paper “Singer on Killing Animals”, Shelly Kagan discusses the question of
justification of killing animals and human beings. First part is devoted to morally permissible
ways of killing animals and in second he discusses those possibilities of killing human beings.
He uses the argumentation of Peter Singer’s Practical Ethics to show the problems in
question and to criticize some of Singer’s points. Kagan starts his analysis with problem of
killing animals with regards to their status as mere animals. He uses the distinction between a
mere animal and human being so he could proceed to the questioning possible situations in
which it is maybe morally permissible to kill a human being. This question leads Kagan to the
assumption of replacement as a possible solution at first to the problem of killing an animal
and then to the problem of using this same method, namely, resolving the issue of killing a
person by replacing it with an identical one. Singer thinks that from the utilitarian point of
view it could be justified to replace human being. In order to avoid this conclusion, he makes
an argument by combining debit model of preference utilitarianism and assumption of the
existence of objective value. Kagan goal is to show that Singer was wrong in making his
argument more complicated, by thinking that preference utilitarianism implies that people are
replaceable. In last part of his paper he tries to show that by holding the view of preference
utilitarianism 1) you do not need to take the debit model and that 2) the preference utilitarian
does not mean that people are necessarily replaceable.

Killing mere animals

Kagan begins by posing a question: is there anything wrong with killing a sentient
creature per se, namely, a sentient animal? His first answer is that killing an animal that would
otherwise have an overall pleasant future prevents it from experiencing the pleasure it would
otherwise get. In defending this view, Singer thinks we cannot hold preference view. Instead,
he believes that we have to adopt a view according to which pleasure has objective value
because sentient being has no present concern for its future pleasure. But Kagan object in
saying that we can hold preference view in on pleasant future of the animal. Namely, if the
pleasure had occurred, the animal would have had a relevant preference. Kagan ascribes to
Singer the debit model of preference view according to which there is no positive value in the
satisfaction of preferences per se but only a negative value or worth in their frustration.

After this, Kagan introduces implication that Singer anylise, the claim that it is wrong
to kill because of the loss of the potential future pleasure. If this is the case, then it is good to
create sentient creature, because it will experience pleasure and will affect total amount of
pleasure. Singer makes a difference between the two views in relation to the implication that
if killing is wrong because of the loss of the potential future pleasure, then it would be good to
create extra sentient beings with pleasant lives. If we advocate “total view“, then we accept
the implication. If we do not accept it then we are representatives of the “prior existence
view”. Kagan thinks that this is wrong reasoning in this place, because the focus should be on
saving the life of a person in question rather than creating someone with pleasant future. So
for Kagan, the right implication should be: if killing someone with a pleasant future is bad,
then saving (lengthen a pleasant life) the life of someone with a pleasant future should be
good.
However, Kagan thinks that we have to address the question of creation new being
when we introduce replaceability. He ascribes the conclusion that merely sentient beings are
replaceable to Singer and the reasoning is that we can imagine a case in which we kill an
animal and replace it with an animal of same kind so the overall amount of pleasure is
preserved. But Kagan object in stating that even if the pleasant life of the new animal does
count as a good, and as great a good as the potential future pleasure that is destroyed by
killing, it doesn’t yet follow that it is morally permissible to kill the first animal. The
underlying assumption is a utilitarian view and that moral permissibility is a simple matter of
adding up the good and the bad consequences. But, we can reject this view from a
deontological point of view, where this overall amount of pleasure is irrelevant, and
replaceability is not permitted.

In analyzing creation of animals, Kagan defines them as 1) inevitables - will exist

regardless of the choice; 2) contingents - existence depends on the particular choice; 3)
actuals - those animals who actually end up existing at some time or the other—past, present,
or future (all inevitables and some contingents) and 4) nonactual contingents - those animals
that could have existed, but never do. So the next question becomes: Which interests are we to
count? There are several views we can take: A) Narrow views count only the interests of the
inevitables; B) Wide views count the interests of contingents as well and C) intermediate
views count the interests of some contingents but not all of them (for example, an
intermediate view might count the interests of all actuals, contingent or inevitable, but not the
interests of nonactual contingents).

Further, we can differentiate from I) symmetrical views and II) asymmetrical views. In
symmetrical views we have Ia) The total view is a wide view. It counts the interests of
everyone who will or might exist. And it does this symmetrically, taking into account both
pleasure and pain; Ib) The prior existence view is narrow, only counting the interests of the
inevitables. But with regard to those interests, it symmetrically counts both pleasure and pain.
And Ic) Actualism which is an intermediate view, counting the interests of all and only
actuals. Thus it counts the interests of all inevitables (since they must be actual) and of all
actual contingents, but it does not count the interests of nonactual contingents. It too treats
pleasure and pain symmetrically, when they count at all. In asymmetrical views we have IIa)
Extreme asymmetry, according to which pain counts against an outcome, but pleasure does
not count in its favor - not even the pleasure of inevitables. IIb) Moderate asymmetry,
according to which both pleasure and pain count for inevitables, but not for contingents. For
contingents, only pain counts, not pleasure. IIc) Impure asymmetry according to which the
interests of all actuals are treated symmetrically (that is, both pleasure and pain count for
actual contingents, and not only for inevitables); but for nonactual contingents only pain
counts. IId) Offsetting asymmetry, according to which pleasure can count to offset pain
(sufficient pleasure can cancel out the reason to avoid a given outcome that would otherwise
be generated by the existence of pain under that outcome) but pleasure can never count in
favor of an outcome (there would, for example, be no reason to promote an outcome with
pleasure but no pain). IIe) Impure offsetting asymmetry according to which pleasure can count
in favor of an outcome, but only in the case of actuals. For nonactual contingents, pleasure
can only be counted to offset pain; it provides no reason in its own right to favor an outcome.

In discussing this views, Kagan find Actualism and the prior existence unacceptable,
while he thinks that total view atlhough deals better with paradoxes, is regarded by many as
implausible because it also implies that we have reason to create extra sentient beings that will
have pleasant lives. Problem of creation of being with miserable life can be solved in holding
a version of asymmetry view. Asymmetry provide reason for cases where there is no reason to
create a being with pleasant or a being with miserable life. Problem with asymmetry views is
that the it treats pleasure in ad hoc manner, it is not clear why pleasure should sometimes fail
to generate a reason in cases where pain nonetheless would. Kagan discuses problems of
asymmetry views in details and conclude that the right view is maybe offsetting asymmetry,
but it too suffers from ad hoc character. At the end of this part of his paper, Kagan chose the
total view as least unacceptable of the various alternatives. On this view it is justified to kill
and animal if you replace it with with a new animal with an equally pleasant life. Conclusion
is that merely sentient beings are morally replaceable.

Killing people

I this part of the paper, Kagan applies the reasoning from the first part and discuses the
question of killing human beings. He asks about the possible reasons why it should be wrong
to kill a human beings. This question leads him to think that reasons could be different from
those against killing an animal. In answering this question, Kagan is discussing about four
reasons that Singer proposed. From preference utilitarian viewpoint 1) killing a person may
affect others who themselves have preferences about their continued existence and 2) killing a
person frustrates his preference to keep on living. From deontological perspective 3) people
have a right to life, grounded in their desire to continue to exist into the future; 4) autonomy -
killing a person interferes with their autonomous choice to go on living. Singer analyzes the
question only within 1) and 2).

Singer thinks that preference utilitarian can not say that it is morally wrong to kill a
person while also creating a new person, with an equally pleasant life, who would not
otherwise exist. If we kill a person and replace it with a person, we will get the same result -
sa satisfied preferences. For preference utilitarian, it seems that people are replaceable. In
answering this problem, Singer adopts debit model of preference satisfaction (there is no posi-
tive value in the satisfaction of a preference, there is only negative value in the existence of
unsatisfied or frustrated preferences.) This way, it is wrong to kill A even though you replaced
it with B because it is only the frustrated preferences of A that counts, satisfied preferences of
B are irrelevant. But debit model have problems serious problems (creation of beings in the
first place are worse scenario than this world) Singer solves by adding the assumption that
pleasure is objective good that can compensate frustration of preferences. This leads to
additional problems for Singer because replacing and lack of satisfied preferences could be
overcomed by having more objective goods in the world. But this leads to additional
problems, because we would have objective value if the person was not killed in the first
place, so the amount of the objective values stays the same but we still have frustrated
preferences. So the results are indeed better if we avoid killing and Singer position is safe. So,
Singer argument is not invalid but complicated. So, Kagan wants to show how Singer could
be simple preference utilitarian and still holds that it is not morally permissible to kill and
replace a human being.

The idea is that we imagine a simple version of preference utilitarianism where For
each frustrated preference we assign a score of -1; and for each satisfied preference we assign
a score of +1. For the sake of argument, human being (H) have 4 stages of life (t1, t2,t3,t4). In
t1, H will have 4 preferences to experience pleasure at t1, t2, t3 and t4, in t2 H will have 3
preferences and so on. if H is left in peace, all of the H’s preferences will be satisfied. The
score would be +10. If the person is killed after t2 than the score would be 3 + -4 = -1. And if
we replace H1 and count only the period from t1 to t4, then the score is -1 + 7 = +6. At the
end, killing a person and replacing him leads to worse results overall. Kagan raise the
objection that we could take into the account all periods, not just from t1 to t4, but he thinks
that this would lead to making the killing obsolete, because letting live and then replacing will
normally have better results than killing and replacing.

Ideje:

Problemi

-Ceo argument se bazira na zelji osobe da nastavi da zivi, ima buduci zivot, itd. Sta je sa
osobama koje ne zele da zive, nemaju trenutno tu zelju ili su iz nekih razloga sprecene da
imaju tu zelju.

-Still, if we are only interested in the value that arises during t1–t4, then we should only look
at the preferences of the replacement that exist during t3 and t4 (disregarding those that exist
later, during t5 and t6). – Zasto bismo uzimali u obzir samo preferencije u t3 i t4?

Kagan sam prikazuje ovaj prigovor:

To be sure, it might be objected that if we take into account all of the results—not just
those that occur during what would have been the first person’s lifetime, but also those that
occur afterward—preference utilitarianism might well sometimes conclude that the results
would be better if we did kill and replace. In principle, after all, the satisfied preferences that
the replacement person will have later in life might well outweigh the frustrated preferences
of the victim. (That won’t be true, as it happens, in our particular example; but it certainly can
be true if the victim is killed late enough in life, and the replacement lives for a long enough
time afterward.)

Ali zakljucuje:

But that still doesn’t mean that the preference utilitarian thinks it permissible to kill and
replace. For if we are going to extend the time frame of our comparison, and take into account
the benefits from replacement due to the increase in the total amount of life that replacement
makes possible (that is to say, take into account the gains from satisfied preferences that occur
after t4), then it is important to bear in mind that replacement will provide even greater
overall gains if it is done without killing. That is, the results would be better still if, instead of
killing first and then replacing, we create the new person only after the first person has lived
out his life. Letting live and then replacing will normally have better results than killing and
replacing. So even when we calculate over the extended time frame, simple killing and
replacement won’t normally be permissible.

Odgovor: